Fish Assemblage Dynamics in the Tagus and Sado Estuaries (Portugal)

Cah. Biol. Mar. (2008) 49 : 23-35

Fish assemblage dynamics in the Tagus and Sado estuaries (Portugal)

Ana NEVES1, Henrique CABRAL1, Ivone FIGUEIREDO2, Vera SEQUEIRA1, Teresa MOURA2 and Leonel Serrano GORDO1 (1) Faculdade de Ciências da Universidade de Lisboa, Departamento de Biologia Animal & Instituto de Oceanografia, Bloco C2, Campo Grande, 1749-016 Lisboa, Portugal Fax: +351 217500028. E-mail: [email protected] (2) Departamento de Recursos Marinhos, IPIMAR, Av. Brasília, 1449-006 Lisboa, Portugal

Abstract: From May 2001 to April 2002 the ichthyofauna from Tagus and Sado estuary was monthly sampled. 25 species from 18 families were identified in Tagus samples, and 38 species from 21 families in Sado estuary. Only 14 species (26%) occurred in the two estuaries, but 14 families (58%) were common for both. Pomatoschistus minutus, Gobius niger, Syngnathus sp and Sardina pilchardus were the dominant species in Tagus estuary. Gobius niger, Monochirus hispidus, Halobatrachus didactylus, Arnoglossus laterna and Diplodus bellottii were the dominant species in Sado. From CCA analysis a separation of Tagus samples from those of Sado is evident. Furthermore, in Sado estuary a spatial pattern is evident, while in Tagus estuary no pattern is evident among sampling areas or seasons. The physical factors which have a major influence are salinity, total organic mater and percentage of mud, being the last two highly related. Comparing the diversity index in the two estuaries the values are higher in Sado estuary probably due to a greater influence of seawater and a higher variability of sediment, which induce to a more rich community.

Résumé : Dynamique des assemblages de poissons dans les estuaires du Tage et du Sado (Portugal). La faune ichthyologique des estuaires du Tage et du Sado a été échantillonnée mensuellement depuis le mois de mai 2001 jusqu’au mois d’avril 2002. On a identifié 25 espèces de 18 familles dans le Tage et 38 espèces de 21 familles dans le Sado. Seules 14 espèces (26%) sont communes aux deux estuaires ; 14 familles (58%) sont communes. Les espèces dominantes du Tage sont Pomatoschistus minutus, Gobius niger, Syngnathus sp et Sardina pilchardus et dans le Sado se détachent Gobius niger, Monochirus hispidus, Halobatrachus didactylus, Arnoglossus laterna et Diplodus bellottii. L’analyse canonique des correspondances a montré une séparation nette entre les échantillons du Tage et ceux du Sado. En outre, une structure spatiale est visible dans l’estuaire du Sado, alors qu’aucune structure, ni spatiale ni saisonnière, n’apparaît pour le Tage. Les facteurs physiques les plus influents sont la salinité, la matière organique totale et le pourcentage de vase, ces deux derniers se trouvant fortement corrélés.

Keywords: Fish assemblage l Environmental factors l Estuarine system l Tagus l Sado l Portugal

Reçu le 2 mai 2007 ; accepté après révision le 11 mars 2008. Received 2 May 2007; accepted in revised form 11 March 2008. 24 FISH ASSEMBLAGES IN TWO ESTUARIES

Introduction Sado estuary has an area of approximately 180 km2 of which 44% is intertidal (Vasconcelos et al., 2007). The river Estuaries have long been known to be important as flow regime presents significant interannual and seasonal nurseries areas, migration routes and habitats for estuarine variations and is highly dependent on the amount of rain residents (Haedrich, 1983). They are also very important occurring in the south part of Portugal, varying from 1 m3 for human activities, and often associated to large urban or s-1 in summer to 60 m3 s-1 in winter (Martins et al., 2001). industrial areas. Along with anthropogenic pressures (e.g. Only 5 small tributaries provide freshwater to the estuary poldering and land-claim, channel management, power and, consequently, the estuary has a high salinity, above stations and water abstraction, barrages) (Cattrijsse et al., 29.5, all over the year (Cunha, 1994). The tidal regime in 2002), estuaries are affected by natural changes in the area is primarily semidiurnal with a range of 0.6 to1.5 environmental condition. Temperature and salinity m in neap tides and of 1.6 to 3.5 m in spring tides (Martins gradients, alterations of water-column turbidity and in et al., 2001; Amaral & Cabral, 2004). chemical composition including changes in nutrients, dissolved gases and trace metals are, among others factors, Sampling procedures prominent features of estuarine systems likely to affect fish From May 2001 to April 2002, 4 sites in the Tagus and 5 populations (Elliott & McLusky, 2002). sites in the Sado estuary were sampled monthly (Fig. 1), Fish have very diverse live history strategies and occupy using a 4-m beam trawl with 10-mm mesh size. This gear is a variety of ecological niches using the estuary in several one of the more appropriate to the estuarine environment different forms. Some species have obligate live history (Potts & Reay, 1987) due to its reduced selectivity when stages in the estuary such as estuarine resident or compared to other gears. Trawls were towed for 15 min at diadromous fishes, others are facultative users (marine or 1.85 km h-1, during daylight and low water of spring tides. freshwater opportunists) and others only occasionally find After capture, specimens were kept at low temperature in their way into an estuary (marine vagrants) (Able, 2005). an icebox and later identified, measured to the nearest mm The rivers Tagus and Sado form two of the major (total length, TL) and weighed to the nearest 0.01 g (wet estuarine systems in the Atlantic coast of the Iberian weight, WW) in the laboratory. The identification was Peninsula (Costa, 1988; Cunha, 1994). Both estuaries conducted according to Whitehead et al. (1986). suffer the influences of industrial complexes, farming, During all samplings, temperature (± 0.1ºC), salinity (± urban areas and fishing (Cabral, 1998), activities that are 0.1) and dissolved oxygen concentration, DO (± 0.1 mg l-1) often linked with alterations of the health status and were measured using a multiparameter water quality probe ecological integrity of estuaries. Over two million people (YSI, model 85). In spring (May), summer (August), live in the area of influence of the Tagus estuary and over autumn (November) and winter (February), sediment 300 000 people live around Sado estuary. samples were collected using a van Veen grab and analysed Because estuaries are water-bodies sensitive to the for granulometric purposes. These samples were washed to adverse effects of human impacts (Elliott & McLusky, remove the fine portion (< 0.063 mm), dried at 70ºC and 2002) it is of major importance to investigate the nature of then sorted in a 1-mm sieve to separate gravel from sand. estuarine communities to aid preservation and management The organic matter in the sediment was also determined by of these areas (Elliott, 2002). the loss of weight on ignition at 450ºC for two hours. The aim of the present study is two-fold: 1) to identify Species were grouped into ecological guilds according to the main patterns of spatial and temporal variation in the common traits of life history, biology and/or behaviour (see fish assemblage of each estuary, and 2) to describe the Elliott & Dewailly, 1995). environmental conditions associated with the fish biota. Data analysis Material and Methods For data analysis, months were grouped in seasons: March, April, May (spring); June, July, August (summer); Study areas September, October, November (autumn); December, Tagus estuary has an area of 325 km2 of which 40% is January, February (winter). k-dominance curves intertidal area (Cabral, 1998) and a tidal range of 4.6 m (Lambshead et al., 1983) were used to graphically account with a mean depth less than 10 m (França et al., 2005). The for diversity outline. For each sampling site, the species mean Tagus river flow is 400 m3 s-1, being highly variable richness (Margalef, 1958), the Shannon-Wiener diversity both seasonally and interannually (França et al., 2005). index (Shannon & Weaver, 1949) and the evenness index Beyond the Tagus, 15 tributaries contribute to the import of (Pielou, 1977) were used to characterize fish temporal and freshwater into the estuary. spatial abundance. Significant differences on the abiotic A. NEVES, H. CABRAL, I. FIGUEIREDO, V. SEQUEIRA, T. MOURA, L. S. GORDO 25

Figure 1. Map of the sampling sites in Tagus and Sado estuaries. Figure 1. Localisation des points d’échantillonnage aux estuaires du Tage et du Sado.

parameters and ecological indices were evaluated among Results sites and seasons (Kruskal-Wallis test) and estuaries (Mann-Whitney test) for a significant level (α) of 0.05 (Zar, Abiotic parameters 1999). Fish (excluding the species with abundance lower than Mean values of temperature and dissolved oxygen were not 1%) and environmental data matrix were submitted to significantly different between estuaries (U = 0.242, p = canonical correspondence analysis (CCA) in order to obtain 0.809, U = 1.691, p = 0.091 respectively). Mean salinity a direct environmental interpretation of the extracted values were significantly higher (always above 30) in Sado ordination axes (Ter Braak, 1994). Despite the fact that estuary (U = -8.7888, p < 0.001), while in Tagus estuary the CCA is a linear technique, it is less susceptible than other values varied between 10.3 and 28.3 (Fig. 2). ordination methods to non-linearities in the data matrix, Seasonally, Tagus estuary presented significant because it explicitly assumes non-linear responses of differences in temperature and salinity (Table 1) varying the species along environmental gradients (McGarigal et al., former from a mean value of 13ºC in winter to 22ºC in 2000). CCA was performed using CANOCO 4.5 (Ter Braak summer. The latter varied from a mean value of 17 in spring & Smilauer, 1998). to 24 in summer. Sado estuary presented significant 26 FISH ASSEMBLAGES IN TWO ESTUARIES

Table 1. Values of the Kruskal-Wallis test for the physical parameters. Significant values (α < 0.05) are signed with *. Tableau 1. Résultats du test de Kruskal-Wallis sur les paramètres physiques. Les valeurs significatives (α < 0,05) sont signalés par *.

Among Seasons Among Sites Tagus Sado Tagus Sado Kruskal-Wallis p value Kruskal-Wallis p value Kruskal-Wallis p value Kruskal-Wallis p value Test H test H test H test H

Temperature 32.9539 < 0.001* 32.6006 < 0.001* 0.3304 0.9542 0.2945 0.9902 Salinity 9.9456 0.0190* 14.5103 0.0023* 19.8265 < 0.001*8.3110 0.0808 Dissolved Oxygen 1.4130 0.7025 13.0051 0.0046* 6.6385 0.0844 2.5963 0.6275

Figure 2. Mean values recorded for temperature, salinity and dissolved oxygen during spring, summer, autumn and winter in the Tagus and Sado estuaries. Figure 2. Valeurs moyennes enregistrées aux estuaires du Tage et du Sado pour la température, la salinité et l’oxygène dissous pen- dant le printemps, l’été, l’automne et l’hiver. A. NEVES, H. CABRAL, I. FIGUEIREDO, V. SEQUEIRA, T. MOURA, L. S. GORDO 27

Figure 3. Sediment composition in the sampling areas of Figure 4. The relative presence of ecological guilds in the Tagus and Sado estuary (gravel - white bar; sand - striped bar; Tagus and Sado estuaries. CA - catadromous species; MA - mud - grey bar). marine adventitious species; MJ - marine juveniles; ER - Figure 3. Composition du sédiment aux stations d’échantil- estuarine residents. lonnage des estuaires du Tage et du Sado (gravier - colonne Figure 4. Présence relative des différents catégories blanche ; sable - colonne rayée ; vase - colonne grise). écologiques aux estuaires du Tage et du Sado (CA - espèces catadromes ; MA - espèces marines occasionnelles ; MJ - juvéniles d’espèces marines ; ER - espèces résidant dans l’estuaire). differences for all the physical parameters (Table 1). Temperature varied from a mean value of 13ºC (winter) to 21ºC (summer), salinity varied from a mean value 32 trawls (Table 2), weighing a total of 40.2 kg. Fourteen (spring) to 35 (summer) and dissolved oxygen from a mean species (29%) from 11 families (44%) occurred in the two -1 -1 value of 4.6 mg l (autumn) to 5.5 mg l (spring) (Fig. 2). estuaries. Eleven species appeared exclusively in the Tagus Spatially, in Tagus estuary, there were significant estuary while 24 were exclusive for Sado estuary (Table 2). differences (Table 1) in salinity values among sites (mean Sand goby Pomatoschistus minutus (Pallas, 1770) values varying from 13.9 in site 1 to 24.5 in site 4). On the dominated the Tagus estuary assemblage with a density contrary no significant differences among sites (Table 1) close to 93%, representing a biomass of 17.5%. The toad- were observed in the temperature and dissolved oxygen fish Halobatrachus didactylus (Schneider, 1801) attained (Fig. 2). In the Sado estuary, no significant differences in the highest biomass (48%) in Tagus estuary. In the Sado the abiotic parameters were recorded (Table 1). estuary, the black goby Gobius niger (Linnaeus, 1758) Sediment composition in the Tagus estuary is principally presented the highest density (22.6%) followed by the mud across all sampling sites, reaching very high values (> Senegal seabream Diplodus bellottii (Steindachner, 1882) 90%) in sites 2 and 3 (Fig. 3). In contrast, sand dominated (14.8%) and scaldfish Arnoglossus laterna (Walbaum, the sediment composition of the Sado estuary sampling 1792) (13.5%). The highest biomass was recorded by D. sites, especially in samples 2, 3 and 4 (> 80% sand). bellotti (19.4%), the sole Solea solea (Linnaeus, 1758) Gravel-sized particles were almost nonexistent in sampling (17.2%) and H. didactylus (17.0 %). sites within the Tagus, but represented a percentage close to According to the definitions in Elliott & McLusky 4% across the Sado estuary (Fig. 3). (2002) only four ecological guilds could be found in both estuaries (Table 3). Freshwater species and the marine Assemblage Structure seasonal migrants were not found in this one-year sample. In the Tagus estuary, 2976 individuals weighing a total of The relative proportion of each guild varied between the 17.7 kg were captured from 48 trawls. These specimens two estuaries (Fig. 4). Resident species were dominant in belonged to 25 species (1 elasmobranch and 24 teleost all aspects (number of species, density and biomass) in species) within 18 families (Table 2). In the Sado estuary, Tagus estuary, P. minutus being the most important species 2310 specimens comprised of 3 elasmobranch (3 families) numerically and H. didactylus the most important in and 35 teleost species (18 families) were caught in 60 biomass. 28 FISH ASSEMBLAGES IN TWO ESTUARIES s D e ( c è y t p i s s e n e – d R ; s E t ; n e s e d i n s i r e r a e m n i s r e a c u è t p s s e e - ’ d R s E e ; n s u e a l . j i o - n d e J a v S M u j ; u s e d e n l t i l r e e a n e n m g o a - i T s J a u c M d c ; o s s e e s i r e i c a n e i u r p t s a s e s m u x s o u e i t a c i t è n n p o s e s e v i . a d – s s a e r i A e r u n a M i o r u ; p t a s s s e e m u m n o – o e d r t a A d b S a o M t d ) a ; 2 n c s - a e s i m e s c c u e 0 è g p 0 p s a 0 s T s 1 e . u e - g o h t A = m n C o i B r : ( ) d n a e o G t s s a E s a ( c a e - e s m u A o y q i i b C b g : d t ) o e e l G n o ) i c E 2 a - ( é t b m d e l i o i r 0 ) u o 0 2 g g - 0 é l 1 t m . a a c d c 0 i n , 0 g i s 0 o e l = 1 c . o è c g D p e ( s = , é e n t i B s o ( s i e t n i s d e s s n d a o o ; p i m t e i o m i r s o i b o c a p u d s t n e m s i a e o c ’ ) l e C 2 p - . s 2 S n m . a u 0 2 d a 0 e e s l l t 0 n 1 b b . e a a d d T T i n i s é = r A. NEVES, H. CABRAL, I. FIGUEIREDO, V. SEQUEIRA, T. MOURA, L. S. GORDO 29

Marine juveniles dominated Sado estuary, D. bellottii being the most representative species in density and biomass. Numerically the resident species were also well represented mainly by G. niger. k-dominance curves for density and biomass (Fig. 5) showed different patterns for Tagus and Sado estuaries. The former showed a high level of dominance mainly in the density, reaching summer and winter 100% at the first species. Sado estuary presented a more even distribution with lower dominance at species level.

Spatial and seasonal pattern Although the number of species caught in the two estuaries is not statistically different (U = -1.152, p = 0.249), the comparison of both the diversity and evenness indices showed significant differences (U = 4.022, p < 0.001, U = 4.443 p < 0.001, respectively). It must be pointed out that Sado estuary presented the higher values in the diversity indices analysed. In Tagus estuary, Shannon- Wiener diversity index and evenness varied significantly (Table 4) along the year (with very low values during summer and autumn and higher records in spring and winter) and among sites (sites 3 and 4 being always more diverse). The species richness was similar along the year but signifi- cantly different among sites (Table 4). In Sado estuary the diversity index and the species richness varied significantly among seasons (Table 4), having the winter the lowest values (Fig. 6). The evenness was statistically different among sites (Table 4). The two first axes of the CCA using both species abundance data grouped according to sampling site and season accounted for 38.0% of the total variance and 84.7% of the relationship between abundance 30 FISH ASSEMBLAGES IN TWO ESTUARIES

Figure 5. k-dominance curves for icthyofauna abundance and biomass in the two estuaries along the four seasons. Figure 5. Courbes de k-dominance pour l’abondance et la biomasse de la faune ichthyologique au cours de l’année dans les deux estuaires.

and the environmental variables. All sampling areas and Table 3. Density (ind.1000 m-2) and biomass (g.1000 m-2) of seasons related to the Sado estuary clustered separated (in fish sampled in Tagus and Sado estuaries grouped by ecological the right edge of the diagram) from those of Tagus estuary guilds. CA- catadromous species; MA - marine adventitious (in the left edge) (Fig. 7). Furthermore, in the former species; MJ - marine juveniles; ER - estuarine residents. estuary, sample areas 1 and 5 were separated from the other -2 -2 Tableau 3. Densité (ind.1000 m ) et biomasse (g.1000 m ) sites, being present in the lower and upper parts of the des poissons échantillonnés aux estuaires du Tage et du Sado diagram respectively while sites 2 and 3 and 4 overlap in agroupés par catégorie écologique (CA - espèces catadromes ; MA - espèces marines occasionnelles ; MJ - jaunes d’espèces the central area. In the Tagus estuary, no pattern is evident marines ; ER - espèces résidant dans l’estuaire). among sampling areas or seasons and all lie in a common group (Fig. 7). The species distributions show 4 assemblages: two Tagus estuary Sado estuary major groups and other two with just a few species present. DB DB A first group enclosing the dragonet Callionymus lyra Linnaeus, 1758 (CLL), reticulated dragonet Callionymus CA 3.10 2731.95 0.18 54.37 reticulatus Valenciennes, 1837 (CLR), D. bellottii (DPB), MA 0.70 165.24 61.05 2338.40 MJ 23.88 864.30 157.80 5908.79 Senegalese sole Solea senegalensis Kaup, 1858 (SLS), ET 2948.89 13917.14 149.86 2581.61 ocellated wedge sole Dicologlossa hexophtalma (Bennett, 1831) (DLH), white seabream Diplodus sargus (Linnaeus, A. NEVES, H. CABRAL, I. FIGUEIREDO, V. SEQUEIRA, T. MOURA, L. S. GORDO 31

Figure 6. Species diversity, evenness and species richness indices variation for sample sites in the two estuaries along the year. Figure 6. Variation des indices de diversité, régularité et richesse spécifique aux stations échantillonnées dans les deux estuaires pendant l’année.

1758) (DPS), short snouted seahorse Hippocampus by low salinity and sediment with both a high level of fine hippocampus (Linnaeus, 1758) (HPH), salema Sarpa salpa particles (Mud) and high organic matter (TOM). (Linnaeus, 1758) (SRS), Thor’s scaldfish Arnoglossus thori The third and fourth groups of species are influenced by Kyle, 1913 (ART), and whiskered sole Monochirus an environment with high salinity and high percentage of hispidus Rafinesque 1814 (MNH) is mainly associated with mud and were present in the Sado estuary. The third group the sample 2 and 3 of the Sado estuary. This group is is more correlated with the sampling area 5, characterized closely related to high salinity (Sal) and with sediment of by the presence of the sand smelt, Atherina presbyter higher grain size particles. Cuvier, 1829 (ATP), the gurnard, Trigla lucerna (Linnaeus, A second group represented by the species European 1758) (TRL) and the European conger Conger conger seabass Dicentrarchus labrax (Linnaeus, 1758) (DCL), (Linnaeus, 1758) (COC). The fourth group encloses the thinlip grey mullet Liza ramada (Risso, 1810) (LZR), anchovy, Engraulis encrasicholus (Linnaeus, 1758) (ENE), pipefish Syngnathus sp (SYS), meagre Argyrosomus regius S. solea (SLL) and A. laterna (ARL) and is strictly (Asso, 1801) (AGR), P. minutus (PMN), common goby associated with sampling area 1. Pomatoschistus microps (Krøyer, 1838) (PMC), sardine The environment parameters dissolved oxygen (DO) and Sardina pilchardus (Walbaum, 1792) (SDP), transparent temperature (Temp), negatively correlated with the second goby Aphia minuta (Risso, 1810) (APM) and European eel axis, have a weaker influence in the ordination when Anguilla anguilla (Linnaeus, 1758) (ANA), is mostly compared with the other variables. associated with the Tagus estuary, being mainly influenced 32 FISH ASSEMBLAGES IN TWO ESTUARIES

Table 4. Values of the Kruskal-Wallis test for the diversity indices. Significant values (α < 0.05) are signed with *. Tableau 4. Résultats du test de Kruskal-Wallis sur les indices de diversité. Les valeurs significatives (α < 0.05) sont signalées par *.

Among Seasons Among Sites Tagus Sado Tagus Sado Kruskal-Wallis p value Kruskal-Wallis p value Kruskal-Wallis p value Kruskal-Wallis p value test H test H test H test H

Shannon index 8.4910 0.0369* 19.7890 < 0.001* 13.9970 0.0029* 2.8133 0.5895 Evenness 14.2630 0.0026* 5.4000 0.1446 13.2980 0.0040* 16.9900 0.0019* Species richness 2.7180 0.4371 23.5100 < 0.001* 19.2620 < 0.001* 3.9390 0.4144 A. NEVES, H. CABRAL, I. FIGUEIREDO, V. SEQUEIRA, T. MOURA, L. S. GORDO 33

Discussion teleost species in her pioneer study on the structure of the fish assemblage of the Tagus estuary. Among the species Both the Tagus and Sado estuary are greatly influenced by that are no longer present or are seldom caught are some anthropogenic activities. Important industrial complexes diadromous species (the allis shad, Alosa alosa (Linnaeus, such as chemical, petrochemical, food processing and 1758), the twaite shad, Alosa fallax (Lacépède, 1803), the smelting are present in both estuaries, mainly in their lower flounder, Platichthys flesus (Linnaeus, 1758)) and some part while, in the upper parts, they are bordered by land typical freshwater species (the carp, Cyprinus carpio used intensively for agriculture and aquaculture (in the case Linnaeus, 1758, and the barb, Barbus bocagei of the Sado estuary). All these activities generally Steindachner, 1864). According to Assis et al. (1992), the contributed to an impoverishment of water quality until the main causes for the impoverishment of the estuarine end of the 1980s. The creation of both the Tagus and Sado community and the staid decline of diadromous species Natural reserves (in 1976 and 1980, respectively) was were the increase of pollution levels (more than 600 associated with the construction of water treatment plants industrial pollution sources were inventoried), over fishing, and the elimination of effluents which have largely dams (due not only to their physical obstruction but also to contributed to an improvement of water quality. the modifications that they produce on estuary hydro- Present conditions in both estuaries are clearly dynamics) and the loss of habitat (and the areas initially favourable to the presence of a diverse fish assemblage. It chosen by fish for reproduction). must be pointed out that, although the area occupied by the Despite their geographical proximity, the Tagus and Sado estuary is almost half of the Tagus estuary, the former Sado estuaries present fish communities with contracting contains a fish assemblage characterized by a higher ecological guilds structure. Marine juvenile species species richness which is, in fact, one of the highest found dominated the Sado estuary fish community in terms of among Portuguese estuaries (Cunha, 1994). Besides high species density, a situation that has been found in a wide productivity, high richness is probably due to the diversity variety of estuarine systems in Europe and worldwide of habitats (oyster beds, mud flats, ellgrass beds) and a (Haedrich, 1983; Elliott & Dewailly, 1995). However, in predominant marine influence of the Sado estuary. In fact, the Tagus estuary, the highest densities were reported for as Whitfield (1994) pointed out, estuaries that constitute resident species, a fact that has been reported to this estuary coastal bays (as Sado estuary) present a higher diversity since the 1980s (Costa, 1982; Cabral, 1998) and common- when compared to other estuarine environments. ly observed in other estuaries along the Portuguese coast The richer fish assemblage present in the Sado estuary (Monteiro, 1989; Rebelo, 1992; Cabral et al., 2001; Gordo was also shown by the diversity indices which expressed & Cabral, 2001). This situation is different from the one the structure of assemblages in terms of species that occurs in northern European estuaries (e.g. the Severn composition and abundance. They have been used in a and Elbe estuaries) where the marine and/or diadromous number of studies of assemblages of estuarine fish species are present in larger numbers than the resident populations (e.g. Costa & Bruxelas, 1989), both as a species. measure of the fluctuation in estuarine populations due to The contribution of the marine juvenile group is high in immigration and emigration and as a measure of the effects both the Sado and Tagus estuaries showing that they of stress, particularly pollution stress, at the assemblage provide nursery grounds for a large variety of species and level. this is true of other Portuguese estuaries (Costa et al., 2002; Comparing the fish assemblage of Tagus estuary with Monteiro, 1989; Rebelo, 1992). However, in a period of 3 the one that was studied during 1980s (Costa, 1982; Costa decades, the pounting, Trisopterus luscus (Linnaeus, 1758), & Bruxelas, 1989), a decrease in its diversity can be P. flesus, T. lucerna and the fivebeard rockling, Ciliata noticed. In fact, Costa (1982) found 3 elasmobranch and 39 mustela (Linnaeus, 1758), that were very abundant at that

Figure 7. Canonical Correspondence Analysis ordination diagram relative to data on fish densities. The first two ordination axes are represented. Open triangles refer to species (codes in the results section); full circles refer to estuary-season-site (estuary: S, Sado; T, Tagus. Seasons: Sp, spring; S, summer; A, autumn; W, winter. Sites: 1 to 5); arrows refer to environmental variables (Temp, temperature; DO, dissolved oxygen; Mud, mud; Sal, salinity; TOM, total organic matter). Figure 7. Diagramme de l’analyse canonique des correspondances selon la densité du poisson. Les deux premiers axes sont représen- tés. Les triangles ouverts désignent les espèces (voir codage à la section des résultats) ; les cercles désignent les estuaires-saisons-points (estuaire : S, Sado ; T, Tage. Saisons : Sp, printemps ; S, été ; A, automne ; W, hiver. Points : 1 à 5) ; les flèches représentent les variables environnementales (Temp, température ; DO, oxygène dissous ; Mud, vase ; Sal, salinité ; TOM, matière organique totale). 34 FISH ASSEMBLAGES IN TWO ESTUARIES time (Costa, 1984; Costa & Bruxelas, 1989) no longer use References the Tagus estuary as a nursery area (Cabral, 1998). The absence of typical freshwater species in both Able K.W. 2005. A reexamination of fish estuarine dependence: estuaries is probably related to the relatively high salinity evidence for connectivity between estuarine and ocean present (particularly in the Sado estuary), preventing their habitats. Estuarine Coastal and Shelf Science, 64: 5-17. presence in the estuaries even in its upper part. Amaral V. & Cabral H.N. 2004. Ecology of the whiskered sole in Canonical correspondence analysis (CCA) has been Sado Estuary, Portugal. Journal of Fish Biology, 64: 460-474. widely used to describe the major patterns of fish Assis C.A., Costa J.L., Costa M.J., Moreira F., Almeida P. & Gonçalves J. 1992. Ameaças à sobrevivência dos peixes assemblages (Godinho & Ferreira, 1998; Castillo-Rivera et migradores do Tejo. Sugestões para a sua conservação. al., 2002) and to evaluate the relationships with Publicações avulsas do INIP, 17: 429-441. environmental factors (e.g. Elliott & Marshall, 2000). Ayvazian S.G., Deegan L.A. & Finn J.T. 1992. Comparison of There were marked differences in the spatial and seasonal habitat use by estuarine fish assemblages in the Acadian and pattern between the Sado and Tagus fish assemblages. The Virginian zoogeographic provinces. Estuaries, 15: 368-383. Tagus estuary assemblage was dominated by the resident Cabral H.N. 1998. Utilização do estuário do Tejo como área de species and the marine juveniles, particularly those that viveiro pelos linguados, Solea solea (L., 1758) e Solea sene- enter the estuary for food and protection (e.g. D. labrax) or galensis Kaup, 1858, e robalo, Dicentrarchus labrax (L., even hatched there (e.g. A. regius). These two groups of 1758). Ph. D. Thesis: Universidade de Lisboa. 396 pp. species have in common the high tolerance to euryhaline Cabral H.N., Costa M.J. & Salgado J.P. 2001. Does the Tagus environments and it is probable that lower salinity in Tagus estuary fish community reflect environmental changes? Climate Research, 18: 119-126. estuary is therefore favourable to their presence. The Sado Castillo-Rivera M., Zavala-Hurtado J. A. & Zárate R. 2002. estuary assemblage was mainly composed of marine Exploration of spatial and temporal patterns of fish diversity juveniles and occasional marine species that are usually and composition in a tropical estuarine system of Mexico. present in the lower zones of estuaries. The elevated Reviews in Fish Biology and Fisheries, 12: 167-177. salinity in the Sado estuary during all the year, favours the Cattrijsse A., Codling I., Conides A., Duhamel S., Gibson R.N., dominance of marine species in the fish assemblage. Hostens K., Mathieson S. & McLusky D.S. 2002. Estuarine Although salinity is the principal abiotic factor that seemed development/habitat restoration and re-creation and their role to discriminate these two fish assemblages, the influence of in estuarine management for the benefit of aquatic resources. other environmental factors (or combination of factors) In: Fishes in estuaries (M. Elliott & K. Hemingway eds), pp. likely play additional roles. Also biotic factors such as 266-321. Blackwell Science: Oxford. Costa M.J. 1982. Contribution à l’étude de l’écologie des pois- density-dependency considered of great importance in sons de l’estuaire du Tage (Portugal). PhD Thesis: Université niche segregation of estuarine fish species (Darnaude et al., Paris VI. 273 pp. 2001) may also play a role in the fish distribution in the Costa M.J. 1988. The Tagus and Mira estuaries (Portugal) and Sado estuary. Furthermore, the variety of seasonal migrants their role as spawning and nursery areas. Journal of Fish from all life history classifications suggests the multiple Biology, 33 (Supplement A): 249-250. roles of Sado estuary in providing shelter and foraging sites Costa M.J. & Bruxelas A. 1989. The structure of fish communi- for juvenile fish as well as a spawning habitat and an ties in the Tagus estuary, Portugal, and its role as a nursery for appropriate physicochemical environment for adult fishes commercial fish species. Scientia Marina, 53: 561-566. (Ayvazian et al., 1992). Costa M.J., Cabral H.N., Drake P., Economou A.N., In conclusion, the Tagus and Sado estuaries are located in Fernandez-Delgado C., Gordo L., Marchand J. & Thiel R. the proximity of large cities and potentially under harmful 2002. Recruitment and production of commercial species in estuaries. In: Fishes in estuaries (M. Elliott & K. Hemingway influence as a result of anthropogenic activities (including eds), pp. 54-123. Blackwell Science: Oxford. dredging). Nevertheless, both estuaries have been able to Cunha P.L. 1994. Estrutura e dinâmica da ictiofauna do estuário maintain very important ichthyological systems that support do Sado. PhD. Thesis: Universidade de Lisboa. 397 pp. resident species, provide a suitable nursery area to marine Darnaude A.M., Harmelin-Vivien M.L. & Salen-Picard C. migratory species and protection for many occasional 2001. Food partitioning among flatfish (Pisces, species, particularly in their juvenile stages. Pleuronectiforms) juveniles in Mediterranean coastal shallow sandy areas. Journal of the marine Biological Association of the United Kingdom, 81:119-127. Acknowledgments Elliott M. 2002. An overview of the status, study and manage- ment of fishes in estuaries. In: Fishes in estuaries (M. Elliott & The work on the Tagus estuary was financially supported K. Hemingway eds), pp. 555-575. Blackwell Science: Oxford. by VALORSUL and the work on the Sado estuary was Elliott M. & Dewailly F. 1995. The structure and components of financed by Fundação para a Ciência e a Tecnologia European fish assemblages. Netherlands Journal of Aquatic (POCTI/MAR/15259/99) Ecology, 29: 397-417. A. NEVES, H. CABRAL, I. FIGUEIREDO, V. SEQUEIRA, T. MOURA, L. S. GORDO 35

Elliott M. & Marshall S. 2000. The biology of fishes in the Monteiro C.C. 1989. La faune ichthyologique de la lagune Ria Humber estuary. In: Coastal Zone Topics: process, ecology and Formosa (sud Portugal). Répartition et organisation spatio- management. 4. The Humber estuary and adjoining Yorkshire temporel des communautés: application à l’aménagement des and Lincolnshire coasts (N. V. Jones & M. Elliott eds), pp. 85- ressources. PhD Thesis: Université des Sciences et Techniques 95. Estuarine and Coastal Sciences Association (ECSA): du Languedoc. 219 pp. Durham. Pielou E.C. 1977. Mathematical ecology. John Wiley & Sons: Elliott M. & McLusky D.S. 2002. The need for definitions in New York. 386 pp. understanding estuaries. Estuarine Coastal and Shelf Science, Potts G.W. & Reay P.J. 1987. Fish. In: Biological surveys of 55: 815-827. costs and estuaries (J. Baker & W. Wolff, eds), pp. 342-373. França S., Vinagre C., Caçador I. & Cabral H.N. 2005. Heavy Cambridge University Press: Cambridge. metal concentrations in sediment, benthic invertebrates and Rebelo J.E. 1992. The ichthyofauna and abiotic hydrological envi- fish in three salt marsh areas subjected to different pollution ronment of the Ria de Aveiro, Portugal. Estuaries, 15: 403-413. loads in the Tagus Estuary (Portugal). Marine Pollution Shannon C.E. & Weaver W. 1949. The mathematical theory of Bulletin, 50: 993-1018. communication. University of Illinois Press: Urbana. 144 pp. Godinho F.N. & Ferreira M.T. 1998. The relative influences of Ter Braak C.J.F. 1994. Canonical community ordination. Part I: exotic species and environmental factors on an Iberian native Basic theory and linear methods. Ecoscience, 1: 127-140. fish community. Environmental Biology of Fishes, 51: 41-51. Ter Braak C.J.F. & Smilauer P. 1998. CANOCO reference Gordo L.S. & Cabral H.N. 2001. The fish assemblage structure manual and user’s guide to Canoco for Windows: software for of a hydrologically altered coastal lagoon: the Óbidos lagoon Canonical Community Ordination (version 4). Microcomputer (Portugal). Hydrobiologia, 459: 125-133. Power: Ithaca, NY. Haedrich R.L. 1983. Estuarine fishes. In: Ecosystems of the world. 26. Estuaries and enclosed seas (B. Ketchum ed), pp. Vasconcelos R.P., Reis-Santos P., Fonseca V., Maia A., Ruano 183-207. Elsevier: Amsterdam. M., França S., Vinagre C., Costa M.J. & Cabral H. 2007. Lambshead P.J.D., Platt H.M. & Shaw K.M. 1983. The detec- Assessing anthropogenic pressures on estuarine fish nurseries tion of differences among assemblages of marine benthic along the Portuguese coast: A multi-metric index and concep- species based on an assessment of dominance and diversity. tual approach. Science of the Total Environment, 374: Journal of Natural History, 17: 859-874. 199–215. Margalef R. 1958. Information theory in ecology. General Whitehead P.J.P., Bauchot M.-L., Hureau J.-C., Nielsen J. & Systems, 3: 36-71. Tortonese E. 1986. Fishes of the North-eastern Atlantic and Martins F., Leitão P., Silva A. & Neves R. 2001. 3D modelling the Mediterranean. Unesco: Paris. 1473 pp. in Sado estuary using a new generic discretization approach. Whitfield A.K. 1994. Fish species diversity in southern African Oceanologica Acta, 24: 1-12. estuarine systems: an evolutionary perspective. Environmental McGarigal K., Cushman S. & Stafford S. 2000. Multivariate Ecology of Fishes, 40: 37-48. Statistics for Wildlife and Ecology Research. Springer-Verlag, Zar J.H. 1999. Biostatistical analysis. 4th ed. Prentice Hall. New New York. 283 pp. Jersey. 663pp.